US5535366A - Method of and circuit arrangement for freeing communications resources, particularly for use by a switching element - Google Patents

Method of and circuit arrangement for freeing communications resources, particularly for use by a switching element Download PDF

Info

Publication number
US5535366A
US5535366A US07/665,788 US66578891A US5535366A US 5535366 A US5535366 A US 5535366A US 66578891 A US66578891 A US 66578891A US 5535366 A US5535366 A US 5535366A
Authority
US
United States
Prior art keywords
memory
time
buffer
free
dedicated
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
US07/665,788
Other languages
English (en)
Inventor
Bodo Pfeiffer
Thomas Banniza
Bozo Cesar
Stefan Wahl
Klaus-Dieter Menk
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alcatel Lucent NV
Original Assignee
Alcatel NV
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Family has litigation
First worldwide family litigation filed litigation Critical https://patents.darts-ip.com/?family=25891029&utm_source=google_patent&utm_medium=platform_link&utm_campaign=public_patent_search&patent=US5535366(A) "Global patent litigation dataset” by Darts-ip is licensed under a Creative Commons Attribution 4.0 International License.
Priority claimed from DE4007747A external-priority patent/DE4007747A1/de
Priority claimed from DE4011357A external-priority patent/DE4011357A1/de
Application filed by Alcatel NV filed Critical Alcatel NV
Assigned to ALCATEL N.V., A CORP OF THE NETHERLANDS reassignment ALCATEL N.V., A CORP OF THE NETHERLANDS ASSIGNMENT OF ASSIGNORS INTEREST. Assignors: BANNIZA, THOMAS, CESAR, BOZO, MENK, KLAUS-DIETER, PFEIFFER, BODO, WAHL, STEFAN
Application granted granted Critical
Publication of US5535366A publication Critical patent/US5535366A/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/38—Information transfer, e.g. on bus
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/54—Store-and-forward switching systems 
    • H04L12/56—Packet switching systems
    • H04L12/5601—Transfer mode dependent, e.g. ATM
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/10—Packet switching elements characterised by the switching fabric construction
    • H04L49/104—Asynchronous transfer mode [ATM] switching fabrics
    • H04L49/105—ATM switching elements
    • H04L49/108—ATM switching elements using shared central buffer

Definitions

  • the present invention relates to a method of managing like units as set forth in the preamble of claim 1, to a circuit arrangement as set forth in the preamble of claim 8, and to a switch element for an ATM system as set forth in the preamble of claim 18.
  • this location will remain blocked for the remaining time of operation. Since the aim is to maintain uninterrupted service and since errors cannot be completely excluded, such a memory will be increasingly blocked unless additional steps are taken.
  • a buffer with a plurality of memory blocks such a whole may be, for example, a multiprocessor system or a multichannel communication system.
  • ATM switch elements One example, with the aid of which the present invention will be explained in more detail, are ATM switch elements.
  • the ATM switch element to be improved in accordance with the invention corresponds to an internal proposal.
  • FIG. 1 is a simplified block diagram of a switch element according to the invention
  • FIGS. 2 to 10 show the writing of cells into the buffer memory and the readout therefrom
  • FIG. 11 shows the occurrence of an error
  • FIG. 12 shows the consequences of the error
  • FIG. 13 shows a circuit for eliminating such errors
  • FIG. 14 shows a supervisory memory, a so-called Time Supervision Field, of an arbiter logic
  • FIGS. 15A and 15B show the row-by-row resetting of the supervisory memory for implementing a self-healing function of the arbiter logic
  • FIG. 16 shows two interconnected arbiter logics together with a common memory area.
  • the switch element shown in FIG. 1 has p input units IP1, . . . , IPp each having one input line I1, . . . , Ip, a multiplexer Mx, a buffer PS having C individually addressable L-bit memory blocks, a demultiplexer Dx, q output units OP1, . . . , OPq, a memory management unit SV, and a routing block W.
  • the input unit IP1 shows the internal structure of the input units in some more detail.
  • the input line I1 runs to a synchronizer SYNC, whose output is connected to a serial-to-parallel converter SPW, and the output of the latter is connected via a unit ZAE for detecting starts and ends of cells to the input end of the multiplexer Mx.
  • the unit ZAE is also connected to one input of the routing block W and to one input of the memory management unit SV.
  • the output unit OP1 shows details of the output units.
  • the output end of the demultiplexer Dx is connected via a unit ZE for detecting ends of cells to a parallel-to-serial converter PSW, whose output is coupled to the output line 01.
  • the output unit further includes an output FIFO OFF and an output register OR.
  • the output FIFO OFF has its input connected to one output of the routing block W, and its output is coupled to the output register OR, which, in turn, is linked bidirectionally with the memory management unit SV.
  • the multiplexer Mx and a network of single lines provide the connection between the input units IP1, . . . , IPp and the central units, namely the buffer PS, the memory management unit SV, and the routing block W.
  • the data stream is converted to an L-bit parallel format.
  • control and clock lines nor the power supply are shown. All these are things for which sufficient solutions are known to the person skilled in the art, i.e., solutions from which he can choose if he knows what information is to be exchanged where and when.
  • switch elements for synchronous systems Such a switch element, which includes a bus system for interconnecting the units, is described in an article by J. M.
  • the memory management unit SV is shown by a table-like representation of its memories.
  • the heart of the memory management unit SV is a list memory LL, which has one location for each of the C memory blocks of the buffer PS. This location has the same address as the associated memory block and can receive the address of another memory block.
  • the list memory LL is empty.
  • the memory management unit SV further includes a memory EFF for storing the addresses of unused memory blocks, which works on the FIFO (first-in-first out) principle and initially contains the addresses of all memory blocks of the buffer PS in a random sequence.
  • the output FIF0s, of which OFF1, OFFi, and OFF16 are shown, are at first empty, too.
  • an "empty" location contains some kind of data, as usual. This may be a value which marks the location as empty, but it may also be a value from the previous use which is no longer valid. In the latter case, steps must be taken to ensure that this value will not be called for any more.
  • the memory management unit SV of the present embodiment further includes C locations of a count memory CC, which can be addressed by the list memory LL.
  • the count memory CC indicates to how many output units the contents of the associated memory block of the buffer PS still have to be outputted.
  • the contents of all locations of the count memory CC are initially 0.
  • the synchronizers of the two input units IP1 and IPk together with the respective serial-to-parallel converters, convert each of the cells into five blocks of equal length.
  • Each of the first blocks SOC1 and SOCk begins with a start-of-cell label SOC.
  • the last blocks EOC1 and EOCk contain an end-of-cell label EOC.
  • the label EOC lies at an arbitrary point within the last block, namely where the cell arriving from outside actually ends.
  • the synchronizer adds blanks to fill the blocks.
  • the remainder of the first blocks, the second to fourth blocks, 1D1, 2D1, 3D1 and 1Dk, 2Dk, 3Dk, respectively, and the portions of the fifth blocks up to the EOC label contain the data of the cells.
  • the input units are cyclically interrogated by the multiplexer Mx.
  • the units ZAE in the input units IP1 and IPk recognize by the SOC labels that information is present which has to be passed on.
  • the address 7 is transferred as the address of an unused memory block into an input register IR1, and the first block SOC1 is stored in the buffer PS in memory block 7.
  • the SOC label goes to the routing block W and to the memory management unit SV.
  • address 7 is noted as the start address of a cell to be outputted by the output unit OPi.
  • a "1" is noted in location 7. The condition shown in FIG. 3 has now been reached.
  • the next block to be transferred into the buffer PS is the block SOCk from the input unit IPk.
  • the memory EFF indicates that this block is to be stored in the buffer in memory block 8.
  • the routing block W determines from the SOC label that this cell is to be placed on both the output line 01 and the output line Oi. Address 8 is therefore stored both in the output FIFO OFF1 and in the output FIFO OFFi, in the latter in the second location behind the 7.
  • the block SOCk itself is stored in the buffer PS in memory block 8, and in the count memory CC, "2" is entered into location 8.
  • the "8" is stored in the input register IRk. The condition shown in FIG. 4 has now been reached.
  • the block 1D1 is stored in the buffer PS in the next unused memory block, 2; in the count memory CC, a "1" is entered into location 2, and in the list memory, a "2" is entered into location 7.
  • Address 7 was temporarily stored in IR1 as the address under which the preceding block of this cell was stored.
  • the other blocks from the input units IP1 and IPk are stored in the same way.
  • the blocks of the cell received from I1 are stored in the memory blocks 7, 2, 3, 4, and 6 in this order, and the blocks of the cell received from Ik are stored in the memory blocks 8, 1, 5, 9, and 13 in this order.
  • the memory EFF indicates only four unused memory blocks in the buffer PS.
  • the list memory LL contains the above-mentioned address sequences, with a special character, here "E", entered instead of a link address for the respective last block.
  • the output FIFO OFF1 indicates that a cell whose first block is stored in the buffer in memory block 8 is to be placed on the output line 01.
  • Address 8 is transferred to the output register OR1 and applied to buffer PS, list memory LL, and count memory CC; the first block is outputted from PS, the link address "1" is transferred from LL to OR1, and the count memory is decremented from “2" to "1", cf. FIG. 6.
  • FIG. 7 shows the next output, which is initiated from OPi.
  • the start address 7 is transferred from OFFi to ORi and applied to PS, LL, and CC; the first block is outputted from PS, the link address 2 is transferred to ORi, and CC is decremented from “1" to "0". This "0" indicates that the contents of block 7 in PS are no longer needed; this block is now free, and its address is returned to the memory EFF for storing the addresses of unused memory blocks.
  • FIG. 8 shows the condition which results when one cell has been outputted to 01 and Oi each.
  • the end of a cell is detected both by the units ZE1 and ZEi for detecting ends of cells and with the aid of the contents of the output registers OR1 and ORi. Units not shown cause dummy blocks to be transmitted.
  • OFFi indicates, however, that a further cell has to be transmitted from the output unit OPi, and that the first block of this cell is stored in PS in memory block 8.
  • the output of the cell begins with the condition in FIG. 9 and ends with the condition in FIG. 10.
  • the buffer PS is designed as a dual-port RAM, so that inputs and outputs can access it independently of each other. This means in particular that the first blocks of a cell can be read out before the last blocks have been written in. This minimizes the delay of the cells and the holding time of the buffer PS.
  • FIG. 11 takes up the situation as it presents itself at the end of the input in FIG. 5. Now, however, it is assumed that because of an error, the link address "1" in the location 8 has been changed to "3" (lightning flash in FIG. 11). This results in various additional errors, which are shown in FIG. 12. Firstly, the cell to which the block in location 8 belongs is outputted falsely, namely twice. Secondly, the entry in the count memory for the falsely outputted blocks becomes negative. However, since only the content "0" has a consequential effect, namely the entry of free memory blocks into the memory EFF, this is of no significance. The blocks with the addresses 1, 5, 9 and 13, which actually should follow the block 8, are not outputted at all (marked with an asterisk in FIG. 12). As a result, they cannot be labelled as free. These blocks are blocked for the remaining time of operation. Since the aim is to maintain uninterrupted service and since errors cannot be completely excluded, such a memory will be increasingly blocked unless additional steps are taken.
  • the starting point for the solution to this problem is the recognition that during error-free operation, the residence time of a cell in the switch element is limited. Data continuously flow off on each output line. The transmission time for each cell is limited. Because of the output FIFOs OFF, the number of cells that can be buffered for each output is limited, too. Thus, a time can be specified after which a just received cell must have been outputted at the latest. The fact that this time possibly cannot be exactly defined but can only be determined from statistical considerations is irrelevant. Hence, the fundamental idea of the solution to this problem is that all entries that are older than this maximum time are regarded as erroneous and are erased.
  • an age memory AT is associated with the list memory LL.
  • the circuit further includes a supervisory circuit US, a comparator V, a first counter Z1, a second counter Z2, and a third counter Z3.
  • the supervisory circuit US can initiate inputs into the age memory AT and the count memory CC, read the contents of the age memory and the count memory, and enter free memory blocks into the memory EFF.
  • the counter Z1 counts at intervals equal to the time required to transmit a cell of average length. At a count corresponding to twice the residence time regarded as a maximum, the counter Z1 is reset. If a block is entered into the buffer PS, the supervisory circuit US causes the instantaneous count of Z1 to be written into the age memory AT by applying the associated address and a write pulse to AT.
  • a memory area particularly a linked-list memory--i.e., a memory with linked address lists--
  • the search for addresses of free memory locations of a memory area is made by means of an arbiter logic which searches a supervisory memory for free memory locations assigned to addresses of the memory area.
  • a supervisory memory also referred to as "Time Supervision Field”
  • FIG. 14 shows a 2-bit field. If one or two bits of a memory are logic "1", the associated location of the memory area is occupied.
  • the arbiter logic searches through the 2-bit field, starting, for example, with the least significant address. As soon as an address of a free location is found, this address is read out by the arbiter logic and marked as occupied by means of a SET signal. The address of the memory area can now be made available for a processing operation.
  • the arbiter logic will find the next free memory location in the supervisory memory very quickly, because occupied locations were marked by means of the SET signal.
  • the lower row of the 2-bit field is set to "0", and after a predetermined time T, the upper row of this memory is also set to "0", as shown in FIG. 15B.
  • the time T must be chosen so that the time between two reset pulses is longer than the period for which the memory location must be made available to the memory area in the worst case.
  • This procedure ensures that the addresses of the supervisory memory are regularly reset to "0", so that free addresses of the memory area are available at regular intervals even if, due to an error, the addresses are indicated as occupied.
  • the supervisory circuit can also be used to initialize the search logic by setting all rows of a supervisory memory, i.e., the two rows of the 2-bit field of FIG. 14, to "0". This results in an especially simple initialization of the circuit.
  • One arbiter logic may be associated with two or more subordinate arbiter logics and determine which of these logics indicate free addresses.
  • the subordinate arbiter logics may be directly assigned to memory areas. It is also possible, however, to assign further subordinate arbiter logics to these subordinate arbiter logics. In this way, a large number of memory locations of a memory area can be managed by relatively few arbiter logics.
  • memory areas of variable size can be managed.
  • a memory area SB is associated with two arbiter logics AL1 and AL2. Associated with both arbiter logics is a common supervisory memory, a Time Supervision Field TSF, with individual address fields of the Time Supervision Field assigned both to the first arbiter logic AL1 and to the second arbiter logic AL2.
  • the searches for free addresses in the individual arbiter logics must be performed in opposite directions.
  • the search for free addresses in the first arbiter logic AL1 for example, is made from left to right, which is indicated by an arrow.
  • the search in the second arbiter logic AL2 is made in the opposite direction, which is also indicated by an arrow.
  • the arbiter logics each begin at an edge region of the memory area SB to look for free addresses. It is quite possible that no problems arise if the addresses of free memory locations of the memory area lie in the field in which the arbiter logics do not overlap. As soon as a free address in the overlap region is read out by one of the arbiter logics, this can be detected by a supervisory logic, which then outputs the respective address of the free memory location in the memory area SB.
  • the associated supervisory memory may form part of the memory area associated with the search logic, so that a regular design of the overall circuit is obtained. As the need for the complex supervisory circuits required with FIFO memories is eliminated, chip area can be saved.
  • simple system initialization is possible by the supervisory circuit, the Time Supervision Control. Errors occurring during a search routine are eliminated by the self-healing function described. Complicated search routines can be dispensed with.
  • arbiter logics are interconnected, large memory areas can be covered. If the arbiter logics overlap as shown in FIG. 16, variable memory areas can be provided, so that optimum memory utilization can be achieved.
  • the search logic for finding the address of a free memory location in a memory area which was explained with the aid of FIGS. 14 to 16, can also be used to find free processors in a multiprocessor system.
  • the individual processors are associated with the supervisory memory.
  • the arbiter logic now searches through the supervisory memory for free processors so as to be able to make the latter available for, e.g., the processing of program steps.
  • the addresses of the processors made available are labeled as occupied in the supervisory memory.
  • the arbiter logic can very easily determine the occupied processors and thus find free processors more quickly.
  • the channels of the data communication system are assigned to the supervisory memory.
  • the arbiter logic finds free channels in the supervisory memory and can make these available for the transmission of data. Such free channels are marked in the supervisory memory and are interpreted as occupied during the next search process. In this way, free channels can be found quickly and simply.
  • the supervisory memory which is associated with the processors or channels, can be reset with the aid of the above-explained Time Supervision Control. In so doing, erroneous bits are set "off" and the supervisory memory is prevented from marking free units as occupied.
  • a counter or one-shot multivibrator is associated with each unit.
  • the counter By setting the counter to a maximum value or by setting the one-shot multivibrator, the beginning of a period of time can be determined.
  • the counter By a clock signal, the counter is reset to zero step by step.
  • a unit with a count of zero or with a reset one-shot multivibrator is then regarded as free. Intentional resetting of the counter or the one-shot multivibrator upon release of the associated unit is not excluded and quite practical.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Debugging And Monitoring (AREA)
US07/665,788 1990-03-12 1991-03-06 Method of and circuit arrangement for freeing communications resources, particularly for use by a switching element Expired - Fee Related US5535366A (en)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
DE4007747.0 1990-03-12
DE4007747A DE4007747A1 (de) 1990-03-12 1990-03-12 Verfahren und anordnung zur speicherplatzverwaltung sowie vermittlungselement
DE4011357.4 1990-04-07
DE4011357A DE4011357A1 (de) 1990-04-07 1990-04-07 Suchlogik-schaltung

Publications (1)

Publication Number Publication Date
US5535366A true US5535366A (en) 1996-07-09

Family

ID=25891029

Family Applications (1)

Application Number Title Priority Date Filing Date
US07/665,788 Expired - Fee Related US5535366A (en) 1990-03-12 1991-03-06 Method of and circuit arrangement for freeing communications resources, particularly for use by a switching element

Country Status (6)

Country Link
US (1) US5535366A (ja)
EP (1) EP0447769A3 (ja)
JP (1) JP3207867B2 (ja)
KR (1) KR100209458B1 (ja)
AU (1) AU642713B2 (ja)
CA (1) CA2037878A1 (ja)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0841783A2 (en) * 1996-08-16 1998-05-13 Nec Corporation ATM switch capable of detecting addressing errors
US5914956A (en) * 1997-02-03 1999-06-22 Williams; Joel R. Cache for improving the connection capacity of a communications switch
US6374331B1 (en) * 1998-12-30 2002-04-16 Hewlett-Packard Company Distributed directory cache coherence multi-processor computer architecture
US6480496B1 (en) * 1997-12-29 2002-11-12 Samsung Electronics Co., Ltd. Apparatus and method for restoring cell storage regions in an ATM switch and method therefor
US6535519B1 (en) * 1998-08-28 2003-03-18 Lsi Logic Corporation Method and apparatus for data sharing between two different blocks in an integrated circuit
US20030072451A1 (en) * 2001-10-16 2003-04-17 Pimentel Roberto J. Method and apparatus for securely transferring wireless data
EP1345359A1 (en) * 2002-03-15 2003-09-17 Broadcom Corporation High speed protocol for interconnecting modular network devices
US20080147764A1 (en) * 2006-07-11 2008-06-19 Bock Alois Motion estimation in image processing systems
US20080155202A1 (en) * 2006-12-26 2008-06-26 Fujitsu Limited Frame Transfer Method and Device
US20090077433A1 (en) * 2004-12-28 2009-03-19 Shearer Robert A Self-healing link sequence counts within a circular buffer

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3647979A (en) * 1970-01-13 1972-03-07 Bell Telephone Labor Inc Program store error detection arrangements for switching systems
US4530054A (en) * 1982-03-03 1985-07-16 Sperry Corporation Processor-addressable timestamp for indicating oldest written-to cache entry not copied back to bulk memory
WO1986003085A1 (en) * 1984-11-14 1986-05-22 American Telephone & Telegraph Company Lockup detection and recovery in a packet switching network
DE3643767A1 (de) * 1986-12-20 1988-07-07 Standard Elektrik Lorenz Ag Speichervermittlungsstelle und verfahren zu deren betrieb
US4947388A (en) * 1988-04-06 1990-08-07 Hitachi, Ltd. Cell switching system of asynchronous transfer mode
US5099475A (en) * 1989-09-14 1992-03-24 Hitachi, Ltd. Switching system
US5130984A (en) * 1990-12-18 1992-07-14 Bell Communications Research, Inc. Large fault tolerant packet switch particularly suited for asynchronous transfer mode (ATM) communication
US5168492A (en) * 1991-04-11 1992-12-01 Northern Telecom Limited Rotating-access ATM-STM packet switch
US5202885A (en) * 1990-03-14 1993-04-13 Alcatel N.V. Atm exchange with copying capability
US5214639A (en) * 1989-08-09 1993-05-25 Alcatel N.V. Communication switching element and method for transmitting variable length cells
US5216669A (en) * 1990-03-23 1993-06-01 Siemens Aktiengesellschaft Method for setting up virtual connections in switching equipment operating according to an asynchronous transfer mode

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4845710A (en) * 1986-12-23 1989-07-04 Oki Electric Industry Co., Ltd. Dynamic buffer supervising system for a data link access protocol control

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3647979A (en) * 1970-01-13 1972-03-07 Bell Telephone Labor Inc Program store error detection arrangements for switching systems
US4530054A (en) * 1982-03-03 1985-07-16 Sperry Corporation Processor-addressable timestamp for indicating oldest written-to cache entry not copied back to bulk memory
WO1986003085A1 (en) * 1984-11-14 1986-05-22 American Telephone & Telegraph Company Lockup detection and recovery in a packet switching network
US4630259A (en) * 1984-11-14 1986-12-16 At&T Bell Laboratories Lockup detection and recovery in a packet switching network
DE3643767A1 (de) * 1986-12-20 1988-07-07 Standard Elektrik Lorenz Ag Speichervermittlungsstelle und verfahren zu deren betrieb
US4947388A (en) * 1988-04-06 1990-08-07 Hitachi, Ltd. Cell switching system of asynchronous transfer mode
US5214639A (en) * 1989-08-09 1993-05-25 Alcatel N.V. Communication switching element and method for transmitting variable length cells
US5099475A (en) * 1989-09-14 1992-03-24 Hitachi, Ltd. Switching system
US5202885A (en) * 1990-03-14 1993-04-13 Alcatel N.V. Atm exchange with copying capability
US5216669A (en) * 1990-03-23 1993-06-01 Siemens Aktiengesellschaft Method for setting up virtual connections in switching equipment operating according to an asynchronous transfer mode
US5130984A (en) * 1990-12-18 1992-07-14 Bell Communications Research, Inc. Large fault tolerant packet switch particularly suited for asynchronous transfer mode (ATM) communication
US5168492A (en) * 1991-04-11 1992-12-01 Northern Telecom Limited Rotating-access ATM-STM packet switch

Non-Patent Citations (10)

* Cited by examiner, † Cited by third party
Title
D. J pel et al: Communications for the Information Age ; IEEE Global Telecom. Conf. & Exhib.; vol. 3; pp. 54.2.1 54.2.7; 28 Nov. 1 Dec. 88. *
D. Japel et al: "Communications for the Information Age"; IEEE Global Telecom. Conf. & Exhib.; vol. 3; pp. 54.2.1-54.2.7; 28 Nov.-1 Dec. '88.
H. M. Deitel; "An Introduction to Operating Systems" Addison-Wesley Publ. Co.; pp. 216-220; 1984.
H. M. Deitel; An Introduction to Operating Systems Addison Wesley Publ. Co.; pp. 216 220; 1984. *
Hj. Keller et al; "A Novel Memory Controller for an ATM Switch" 1990 Intern'l Zurich Sem. on Digital Commun.; pp. 103-114; 5-8 Mar. '90.
Hj. Keller et al; A Novel Memory Controller for an ATM Switch 1990 Intern l Zurich Sem. on Digital Commun.; pp. 103 114; 5 8 Mar. 90. *
J. M. Cotton, et al., "ITT 1240 Digital Exchange Digital Switchin Network", Electrical Communication, vol. 56, No. 2/3, 1981.
J. M. Cotton, et al., ITT 1240 Digital Exchange Digital Switchin Network , Electrical Communication, vol. 56, No. 2/3, 1981. *
S L Min et al; A Timestamp based Cache Coherence Scheme Proc. 1989 Intern l Conf. on Parallel Process g; vol. 1, pp. I23 I32; 8 Aug. 89. *
S-L Min et al; "A Timestamp-based Cache Coherence Scheme" Proc. 1989 Intern'l Conf. on Parallel Process'g; vol. 1, pp. I23-I32; 8 Aug. '89.

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0841783A2 (en) * 1996-08-16 1998-05-13 Nec Corporation ATM switch capable of detecting addressing errors
EP0841783A3 (en) * 1996-08-16 1999-01-13 Nec Corporation ATM switch capable of detecting addressing errors
US6034959A (en) * 1996-08-16 2000-03-07 Nec Corporation ATM switch capable of detecting addressing errors
US5914956A (en) * 1997-02-03 1999-06-22 Williams; Joel R. Cache for improving the connection capacity of a communications switch
US6480496B1 (en) * 1997-12-29 2002-11-12 Samsung Electronics Co., Ltd. Apparatus and method for restoring cell storage regions in an ATM switch and method therefor
US6535519B1 (en) * 1998-08-28 2003-03-18 Lsi Logic Corporation Method and apparatus for data sharing between two different blocks in an integrated circuit
US6374331B1 (en) * 1998-12-30 2002-04-16 Hewlett-Packard Company Distributed directory cache coherence multi-processor computer architecture
US20030072451A1 (en) * 2001-10-16 2003-04-17 Pimentel Roberto J. Method and apparatus for securely transferring wireless data
EP1345359A1 (en) * 2002-03-15 2003-09-17 Broadcom Corporation High speed protocol for interconnecting modular network devices
US20030174719A1 (en) * 2002-03-15 2003-09-18 Broadcom Corporation High speed protocol for interconnecting modular network devices
US7161948B2 (en) 2002-03-15 2007-01-09 Broadcom Corporation High speed protocol for interconnecting modular network devices
US20090077433A1 (en) * 2004-12-28 2009-03-19 Shearer Robert A Self-healing link sequence counts within a circular buffer
US7978705B2 (en) * 2004-12-28 2011-07-12 International Business Machines Corporation Self-healing link sequence counts within a circular buffer
US20080147764A1 (en) * 2006-07-11 2008-06-19 Bock Alois Motion estimation in image processing systems
US20080155202A1 (en) * 2006-12-26 2008-06-26 Fujitsu Limited Frame Transfer Method and Device
US8065450B2 (en) * 2006-12-26 2011-11-22 Fujitsu Limited Frame transfer method and device

Also Published As

Publication number Publication date
EP0447769A3 (en) 1994-12-21
AU642713B2 (en) 1993-10-28
EP0447769A2 (de) 1991-09-25
JPH04220837A (ja) 1992-08-11
AU7260491A (en) 1991-12-05
KR100209458B1 (ko) 1999-07-15
JP3207867B2 (ja) 2001-09-10
KR910017299A (ko) 1991-11-05
CA2037878A1 (en) 1991-09-13

Similar Documents

Publication Publication Date Title
EP0581486B1 (en) High bandwidth packet switch
US4380063A (en) Flow control mechanism for block switching nodes
US3979733A (en) Digital data communications system packet switch
EP0531599B1 (en) Configurable gigabit/s switch adapter
US4748617A (en) Very high-speed digital data bus
EP0054077B1 (en) Method of transmitting information between stations attached to a unidirectional transmission ring
EP0276349B1 (en) Apparatus for switching information between channels for synchronous information traffic and asynchronous data packets
US5081622A (en) Method and apparatus for distributed queue multiple access in a communication system
CA2058816C (en) Common memory switch for routing data signals
US6411599B1 (en) Fault tolerant switching architecture
US5592476A (en) Asynchronous transfer mode switch with multicasting ability
US5210750A (en) Method and apparatus for distributed queue multiple access in a communication system
EP1950932A1 (en) System for transmitting data within a network between nodes of the network and flow control process for transmitting said data
JPH07321823A (ja) マルチキャスティング機能を備えた装置
US5319360A (en) ATM exchange which resequencies the cells using their sequence numbers
US5173898A (en) Multiple-access control for a communication system with order pad passing
US5535366A (en) Method of and circuit arrangement for freeing communications resources, particularly for use by a switching element
US5164937A (en) Packet concentrator and packet switching system
JPH02252341A (ja) ディジタル信号の統計的多重化方法
KR0161753B1 (ko) 비동기 전달 모드 셀 다중화 제어장치
KR960003225B1 (ko) 서비스 품질(qos)등급에 따른 atm 셀 다중화 처리 장치
EP0614585B1 (en) Method and unit for the reconstruction of the correct sequence of atm cells
EP0477242A1 (en) DATA TRANSFER NODE.
Blauman Labeled slot multiplexing: a technique for a high speed, fiber optic based, loop network
EP1914944A2 (en) ATM cell data transmission control

Legal Events

Date Code Title Description
AS Assignment

Owner name: ALCATEL N.V., STRAWINSKYLAAN 341, 1077 XX AMSTERDA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNORS:PFEIFFER, BODO;BANNIZA, THOMAS;CESAR, BOZO;AND OTHERS;REEL/FRAME:005631/0737

Effective date: 19900903

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

FPAY Fee payment

Year of fee payment: 4

FPAY Fee payment

Year of fee payment: 8

REMI Maintenance fee reminder mailed
REMI Maintenance fee reminder mailed
LAPS Lapse for failure to pay maintenance fees
STCH Information on status: patent discontinuation

Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362

FP Lapsed due to failure to pay maintenance fee

Effective date: 20080709